https://github.com/bhklab/nnunet_segmentation_model_benchmark
https://github.com/bhklab/nnunet_segmentation_model_benchmark
Last synced: 3 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/bhklab/nnunet_segmentation_model_benchmark
- Owner: bhklab
- Created: 2022-08-04T22:03:37.000Z (almost 3 years ago)
- Default Branch: main
- Last Pushed: 2022-08-11T21:26:43.000Z (almost 3 years ago)
- Last Synced: 2024-12-30T03:22:53.952Z (5 months ago)
- Language: Jupyter Notebook
- Size: 8.44 MB
- Stars: 0
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# nnUnet_segmentation_model_benchmark
The data that was used for this project is stored at '/cluster/projects/radiomics/Gyn_Autosegmentation/Task80_gyn' in H4H radiomics folder.
## How Testing Robusness of a Segmentation Model Works
### 1.Generate a Benchmarking Dataset
>- Change the input and output path in /segmentation_robustness/examples/dataset_generation.py and run this file.
>- The transformations for benchmarking dataset generation are stored in /segmentation_robustness/examples/roodmri/transforms/defaults.py, which can be modified if needed.
### 2.Testing the Model on the Benchmarking Dataset
>- Instructions for running nnUnet model inference can be found at https://github.com/MIC-DKFZ/nnUNet#run-inference.
>- Other models such as UNet/UNetR can be tested with the testing file segmentation_robustness/test.py.
### 3.Calculating Evaluation Metrics
>- Evaluation metrics for each severity level + transformation combination can be calculated using /segmentation_robustness/generate_metrics.py, there are different functions in this file for multilabel/siglelabel segmentation.
>- A csv file will be generated from generate_metrics.py.
### 4.Calculating Degradation Metrics
>- Be sure to add evaluation metrics for the clean image to the csv file generated form step 3.
>- Degradation Metrics can be calculated using /segmentation_robustness/examples/metric_calculations.py, column names should be change accordingly.
>- Please see https://arxiv.org/abs/2203.06060 for formula and interpretations of the degradation metrics.